Blog
Assorted updates and insights from the LJ team.
Articles
41 articles
Your Next Breach Won't Come Through the Contact Form
The First Agent Your Site Should Serve Is Your Own
We Built Our Chat Agent the Way Our Research Told Us To
Three Files Nobody Reads
The Index Goes in the Context
One Sentence Beats Every File
Nobody Reads llms.txt (Yet)
Introducing AEO Bench: Measuring Agent Readiness Instead of Guessing
The Cheapest Model Passed Every Gate
What It Means to Make Your Website Agent-Ready
Give the Model a Tool, Not a Rule
The $effect Trap: Why Svelte's Escape Hatch Gets Overused
The Error Message Didn't Matter
It Asked About Things It Already Knew
The Twenty-First Note: How Three Neutral Components Compose into Silent Data Loss in LLM Memos
Repeating the Goal Doesn't Make It Yours
Only the Frontier Knows When to Ask
The Frontier Doesn't Need the Training Wheels (We Re-Ran Everything to Find Out)
The Model Always Knew What It Couldn't See
Stronger LLMs Follow Conflicting Instructions More Literally, Not Less
Undo That: The One-Line Echo That Replaces a Transcript
Hand It Everything It Needs: 23 Pre-Registered Studies on LLM Document Editing
Who Writes the Memo? Auditing the Safety Net We Shipped
Views Carry Values, Memos Carry Goals: Our First Judge-Graded Study
The Two Things Your Agent Can't See: Memos and Mentioned Nodes
The Thirty-Sixth Edit: Long LLM Sessions Don't Need Memory Either
barkup 0.5: Your Code Finds the Targets Now
Your Agent Doesn't Need a Memory: Two Worked Examples Replace Session History
barkup 0.4: The Model Finds the Node Now
Then We Found the Cheap Part: One Search Call Grounds LLM Tree Edits
We Tried to Delete the Hard Parts. The Benchmark Said No.
Stable IDs Are All You Need: Seven Studies on Letting LLMs Edit Trees
barkup 0.3: Focused Views, or Why the Model Doesn't Need to See Your Tree
Your Agent's Session Is Drifting (and the Fix Is Cheaper Than the Bug)
The Model Doesn't Need to See Your Tree
We Found the Crossover (It Wasn't Where Anyone Looked)
A Deprecated Accessor That Still Typechecks Broke My Benchmark (and Maybe Your Agent)
barkup 0.2: We Shipped What the Benchmark Told Us
We Benchmarked It: What Held Up in 'HTML as a Native Data Format for LLMs', and What Didn't
Content Is an Overlay: Separating Words from Structure in an AI Document Editor
